Sixth Assessment Report of the Intergovernmental Panel on 
Climate Change (IPCC-AR6;[Fox-Kemper et al. 2021]). The 
IPCC ARS6 vertical land motion (VLM) component is based on 
historic tide gauge trends, which then are extrapolated into 
the future. This approach inherits the spatial heterogeneity of 
the observational network, producing a VLM field that is not 
fully smooth and may contain minor discontinuities, particu- 
larly in areas where land motion varies markedly over short 
distances. The report itself states that ‘there is low to me- 
dium confidence in the GIA and VLM projections employed 
in this Report. In many regions, higher-fidelity projections 
would require more detailed regional analysis’ (Fox-Kemper 
et al. 2021). For regions such as the North Sea-Baltic Sea do- 
main, where GIA-driven uplift gradients are steep, a region- 
ally optimised model provides a more coherent and internally 
consistent representation of VLM. 
We created an optimised, region-specific set of sea level rise 
projections for the North Sea and Baltic Sea by combining two 
existing datasets: VLM data from a semi-empirical model by the 
Nordic Geodetic Commission (NKG) by Vestel et al. (2019) and 
the IPCC-AR6 projections of absolute sea level rise without the 
VLM component (‘novlm’) (Garner et al. 2021; Kopp et al. 2023). 
The regional land uplift model NKG2016LU was selected over 
other GIA models with a global application, such as ICE-6G or 
ICE-7G_NA, since it has a high-resolution and incorporates a 
large number of observations and also includes a geophysical 
GIA model. NKG2016LU is locally calibrated specifically for 
Fennoscandia and the Baltic region. Such locally adjusted infor- 
mation is essential for safeguarding coastal infrastructure, en- 
suring the resilience of transportation routes, and enhancing the 
management of coastal defences like dikes (Hinkel et al. 2018; 
Marijnissen et al. 2020; Meier et al. 2022). 
Similar adjustments to sea level projections have already 
been made for several other regions, such as the Northern 
Mediterranean Coasts (Vecchio et al. 2024), the Netherlands 
(Vermeersen et al. 2018) as well as Denmark (Su et al. 2021), 
which all improve the accuracy of local sea level rise projections 
by better accounting for vertical land motion. 
2 | Data Description and Development 
2.1 | Input Data 
2.1.1 | IPCC AR6 
The dataset IPCC AR6 WGI Sea Level Projections’ (Garner 
et al. 2021) provides sea level rise projections developed for 
IPCC-AR6. It includes detailed estimates of global and regional sea 
level changes under various greenhouse gas emission scenarios. 
The dataset (hereafter ‘IPCC’) encompasses contributions from 
different sources that is thermal expansion, melting of glaciers 
and ice sheets, changes in terrestrial water storage, and vertical 
land motion. An additional dataset similar but excluding only the 
vertical land motion is also provided (hereafter IPCC novlm’). 
Projections span from the historical period up to the year 2100, 
with some extended simulations reaching beyond 2100 up to 2150 
on a 1X1 grid. The dataset also includes associated uncertainties 
and is currently the most comprehensive database for researchers, 
„l 
policymakers and planners to understand potential future sea 
level changes and to develop adaptive strategies for mitigating the 
impacts of sea level rise on coastal communities and ecosystems. 
2.1.2 | NKG2016LU 
We utilise vertical land motion rates derived from the official land 
uplift model NKG2016LU of the NKG, a semi-empirical model fo- 
cusing on land uplift in the Fennoscandian region, as detailed by 
Vestol et al. (2019). This model was developed within the Working 
Group of Geoid and Height Systems of the NKG. It combines an 
empirical model with a geophysical model. The empirical model 
incorporates geodetic data like levelling and time series of Global 
Navigation Satellite System (GNSS) data, whereas the geophysical 
model of GIA supplements data in regions with limited observa- 
tions. Uncertainty in the model results from both the observational 
data and the GIA model, which are combined to provide a compre- 
hensive estimate. The underlying GNSS time series covers the time 
period 2000-2014, and the resulting uplift therefore represents the 
average uplift for that time period. The uplift data are referenced 
relative to the geoid (NKG2016LU_lev’) and give a constant rate of 
uplift and uncertainty for each grid cell on a 1/6Xx1/12° grid. For 
the purpose of this study, we assume the uplift rates to be constant 
in time. 
2.2 | Methods 
2.2.1 | Calculating Optimised Regional Sea Level Rise 
(RSLR) Projections 
1. Sea level change independent of land uplift: In order to ob- 
tain a sea level change without the impact of land uplift, 
we used the dataset provided by Garner et al. (2021), which 
excludes only the ‘vertical land motion’ data (SLCypcc novim) 
and provides uncertainties in quantiles ranging from 0% to 
100%. 
Interpolation of IPCC Data to NKG Grid: To achieve com- 
patibility of the IPCC data with the NKG model, we biline- 
arly interpolated the IPCC data (after preprocessing, see * 
below) onto the NKG grid. Interpolating the coarser field 
(SLCjpcc,novim) Onto a finer grid (NKG) was chosen because 
its smooth, large-scale variations can be accurately repre- 
sented at a finer resolution without introducing inconsist- 
encies. This allows localised details and regional variations 
from the finer dataset to be incorporated in the final data- 
set while ensuring consistency with the global field. 
Addition of NKG VLM Median: Using the NKG2016 model, 
which provides a constant rate of land uplift, we calculated 
cumulative land uplift values per grid cell and per decade. 
In this context, negative uplift represents sea level rise. 
Finally, we added the extrapolated data (- LUygg) to IPCC 
novlm sea level change values (SLCypcc,novm)- It Should be 
noted that the constant VLM from the NKG2016LU model 
was added uniformly to all quantiles in the IPCC AR6 pro- 
jections. No error propagation was applied. This calcula- 
tion is represented by Figure 1 and the following equation: 
DASNordicSLR(g, t) = SLCrpccnovum(Q; D + — LUNG 
Geoscience Data Journal, 2026